\Quarks-2020", Quantum Gravity and Cosmology (dedicated to A.D. Sakharov's centennial) Renormalization group inspired autonomous equations for secular effects in de Sitter space A.Yu. Kamenshchik University of Bologna and INFN, Bologna L.D. Landau Institute for Theoretical Physics, Moscow June 4{8, 2021 Based on: Alexander Yu. Kamenshchik and Tereza Vardanyan∗, Renormalization group inspired autonomous equations for secular effects in de Sitter space, Physical Review D 102 (2020) 065010 ∗Dipartimento di Fisica e Chimica, Universit`adi L'Aquila and INFN, Laboratori Nazionali del Gran Sasso, Italy Content 1. Introduction and Motivations 2. Massless scalar field in de Sitter Space in Flat Coordinates 3. Autonomous equations 4. Application of the autonomous equations to the problem of secular growth 5. Comparizon with the stochastic approach 6. Schwinger{Keldysh technique 7. Yang-Feldman equation for a long-wave part of a scalar field on de Sitter background 8. Conclusion Introduction and Motivations I The de Sitter universe is a spacetime with positive constant 4-curvature that is homogeneous and isotropic in both space and time. I It is completely characterized by only one constant and has as many symmetries as the flat (Minkowski) spacetime. I De Sitter universe plays a central role in understanding the properties of cosmological inflation. I Inflation is a stage of accelerated expansion of the early Universe. The expansion is quasi-exponential, and at lowest order it can be approximated by de Sitter space. I The inflationary stage allows the growth of quantum fluctuations, which are necessary to explain observed large-scale structure of the Universe. So it is important to study quantum field theory in de Sitter background. Introduction and Motivations I It has been known for some time that perturbatively calculated correlation functions of certain quantum field theories set in an expanding background grow secularly (infinitely) with time. I This growth can lead to a breakdown of the perturbation theory past a certain point in time. I The renormalization group method is very effective in quantum field theory. I Dynamical renormalization group permits to improve perturbative solutions of some complicated differential equations. I The attempts to apply the dynamical renormalization group to the treatment of secular effects in de Sitter spacetime did not reproduce known results. Introduction and Motivations I In this work we considered a massless scalar field with quartic self-interaction in de Sitter spacetime and developed a semi-heuristic method for taking the late-time limit of a series of secularly growing terms obtained from quantum perturbative calculations. I We compared our results with the well-known stochastic approach (Starobinsky, 1986, Starobinsky and Yokoyama, 1994). De Sitter Space in Flat Coordinates We consider the de Sitter spacetime represented as an expanding spatially flat homogeneous and isotropic universe with the following metric ds2 = dt2 − a2(t)(dx 2 + dy 2 + dz 2) ; where the scale factor a(t) is a(t) = eHt ; −∞ < t < 1; and H is the Hubble constant that characterizes the rate of expansion. If we introduce a conformal time coordinate, given by η(t) ≡ dt a−1(t) R ds2 = a2(η)(dη2 − dx 2 − dy 2 − dz 2); 1 a(η) = − ; −∞ < η < 0: Hη Physical distances: `phys = a(η)` = −`=(Hη). Physical energy or momentum: kphys = k=a(η) = −kHη. Massless Scalar Field in de Sitter: Equation of Motion, Basis Functions and Quantization p 1 S = d 4x −g g µν@ φ∂ φ − V (φ) 2 µ ν Z We consider λ V (φ) = φ4; λ 1: 4 The equation of motion for the free theory (λ = 0) is r2 φ¨(~x; t) + 3Hφ_(~x; t) − φ(~x; t) = 0 : a2 Transitioning to conformal time and making Fourier transformation d 3~k φ(~x; t) = φ (η)ei~k·~x ; (2π)3 k Z 2 φ00(η) − φ0 (η) + k2φ (η) = 0; k = j~kj: k η k k Equation of Motion, Basis Functions and Quantization General solution of the Klein-Gordon equation −ikη ikη φk (η) = C1(1 + ikη)e + C2(1 − ikη)e : Now, φ can be decomposed as 3~ d k i~k·~x ∗ −i~k·~x y φ(~x; t) = u (η)e a~ + u (η)e a ; (2π)3 k k k ~k Z n o where a~k j0i = 0; and uk (η) ∼ φk (η): How to choose C1 and C2 ? For modes with large momentum, kphys = (−kHη) H ! short physical wavelength; the theory should behave like in flat spacetime, hence iH −ikη uk (η) = p (1 + ikη)e : 2k3 Such a choice is called the Bunch-Davies vacuum. Expectation Values of the Free Theory (λ = 0) and Secular Growth d 3~k h0jφ(~x; t)φ(~y; t)j0i = u (η)u∗(η)ei~k·(~x−~y) (2π)3 k k Z H2 d 3~k (1 + k2η2) = ei~k·(~x−~y) : (1) 2 (2π)3 k3 Z We would like to find the late-time behavior of the long-wavelength part of (1), i.e., the part coming from the modes with physical momenta much less than the Hubble scale, −kHη H. In the case of coinciding spatial points (~x = ~y) H2 −1/η dk h0jφ2(~x; t)j0i = (1 + k2η2) L 4π2 k Zκ H2 1 κ2η2 = − ln (−κη) − + ; (2) 4π2 2 2 where we introduced an infrared cutoff κ for the comoving momentum k, since the integral is divergent at k = 0. Expectation Values of the Free Theory (λ = 0) and Secular Growth For t ! 1 (i.e., −κη 1), the first term in (2) dominates, so in the late-time limit we have H3 h0jφ2(~x; t)j0i = (t − t ) ; L 4π2 0 where t0 ≡ (1=H) ln(κ/H); thus, it grows linearly with time. I This time-dependence is a sign of breakdown of de Sitter invariance. 1 I It was shown by B. Allen (1985) , and subsequently by B. Allen and A. Folacci (1987)2 that no de Sitter invariant state exists for a massless scalar field. 1B. Allen, Phys. Rev. D 32 (1985), 3136 2B. Allen and A. Folacci, Phys. Rev. D 35, 3771 (1987) Expectation Values and Secular Growth λ 4 In the presence of quartic interaction, V (φ) = 4 φ , the 2 leading late-time behavior of hφ (~x; t)iL can be calculated using perturbation theory and the \in-in" (Schwinger - Keldysh) formalism H3 H5 H7 hφ2(~x; t)i = t − λ t3 + λ2 t5 + O(λ3): (3) L 4π2 24π4 80π6 I The perturbative calculation of the long-wavelength part of hφ2(~x; t)i gives a series with terms that behave like λn(Ht)2n+1. p I When Ht > 1= λ, the perturbation theory breaks down, so it can't make reliable predictions at late times. I It is disturbing that expectation values like (3) do not approach a finite limit, since they can enter in the expressions for energy density and observable correlation functions. I There are also subdominant secular terms present: they are suppressed by additional powers of λ with respect to the leading terms, so we ignore them here. Autonomous Equations I We would like to calculate the expectation value of a product of field operators. I We do not have the dynamical equation governing this function, but we have some information obtained by perturbative methods. For λ 1, f (t) = At − λBt3 + O(λ2) ; (4) where A and B are some known positive constants. I As t grows, the perturbation theory breaks down and the expansion (4) can no longer be trusted. I From (4), f (t) ! 1 as t ! 1. I But we can guess from physical considerations that f (t) should be finite. I How can we model the correct behavior and follow what happens at late times? Autonomous Equations Our proposal: construct a simple autonomous first-order differential equation df = F [f (t)] dt that produces the first two terms of the expression (4) by iterations. Denote y(t) ≡ A(t − t0), then f (t) = At − λBt3 + O(λ2) ! B f (t) = y(t) − λ [y(t)]3 + O(λ2) : (5) A3 Differentiating (5), df 3B = A − λ y 2 + O(λ2) : dt A2 Within the given accuracy, y 2 on the right side of this equation can be replaced by f 2, df 3B = A − λ f 2: (6) dt A2 The iterative solution of (6) gives us the original expansion. Autonomous Equations: λ-order But it's also possible to solve the equation df 3B = A − λ f 2 dt A2 explicitly. The solution, with initial condition f (0) = 0, is A3 3λB f (t) = tanh (t) : r3λB "r A # The remarkable feature of this expression is that it is regular for all values of t, and when t ! 1, one has A3 f (t) ! : r3λB Autonomous Equations: λ2-order In principle, this procedure can be generalized for the situation when we have more than two terms coming from perturbation theory f (t) = At − λBt3 + λ2Ct5 + O(λ3) : The corresponding autonomous equation is df 3B 6B2 5AC = A − λ f 2 + λ2 − 1 f 4 : (7) dt A2 A5 6B2 This equation is also integrable and we can obtain its implicit solution t = t(f ). In general it is not possible to find the explicit form of f (t). But in some cases we can obtain the solution of Eq. (7) in the form of perturbative expansion in a small parameter. Autonomous Equations: λ2-order The corresponding autonomous equation is df 3B 6B2 5AC = A − λ f 2 + λ2 − 1 f 4 : dt A2 A5 6B2 5AC If 6B2 − 1 = 0, we are back to the previous equation and its solution.